In recent years, the implementation of autonomous vehicles has been widely discussed worldwide. In particular, urban transportation demand is expected to change significantly when autonomous taxis (ATs) are introduced. Thus, planners must anticipate changes in traffic conditions and the number of users of other transport modes. Therefore, changes in travel behavior and traffic conditions must be quantified with respect to comprehensive changes in the service level of ATs, including changes in fares and the possibility of delays. However, previous studies have not sufficiently considered factors such as the intention to use ATs, the interrelationship between mode choice and traffic congestion, and the impact of ATs on public transit. Therefore, they are not applicable to city-level transportation demand forecasts. The purpose of this study was to propose an easily implementable method for forecasting urban transportation demand when AT services are adopted, which overcomes such problems. In this study, we developed a combined modal split-assignment model and analyzed the effects of various AT service implementation scenarios in Nagoya, Japan. The results showed that the number of trips using taxis will increase by a factor of 11 when AT services are implemented. Additionally, the usage of other modes is expected to decrease by 4%–5%, except for rail usage with a decrease of 1.5%.

近年来，自动驾驶汽车的实施已在全球范围内广泛讨论。特别是，随着自动驾驶出租车 (AT) 的推出，城市交通需求预计将发生显着变化。因此，规划者必须预测交通状况的变化和其他交通方式的用户数量。因此，必须根据 AT 服务水平的综合变化（包括票价变化和延误可能性）来量化出行行为和交通状况的变化。然而，之前的研究没有充分考虑诸如使用 AT 的意图、模式选择与交通拥堵之间的相互关系以及 AT 对公共交通的影响等因素。因此，它们不适用于城市级的交通需求预测。本研究的目的是提出一种易于实施的方法，用于在采用 AT 服务时预测城市交通需求，从而克服这些问题。在这项研究中，我们开发了一种组合模式拆分分配模型，并分析了日本名古屋各种 AT 服务实施方案的影响。结果显示，在实施 AT 服务后，使用出租车的出行次数将增加 11 倍。此外，除铁路使用量下降 1.5% 外，其他交通方式的使用量预计将下降 4%–5%。我们开发了一个组合模式拆分分配模型，并分析了日本名古屋各种 AT 服务实施方案的影响。结果显示，在实施 AT 服务后，使用出租车的出行次数将增加 11 倍。此外，除铁路使用量下降 1.5% 外，其他交通方式的使用量预计将下降 4%–5%。我们开发了一个组合模式拆分分配模型，并分析了日本名古屋各种 AT 服务实施方案的影响。结果显示，在实施 AT 服务后，使用出租车的出行次数将增加 11 倍。此外，除铁路使用量下降 1.5% 外，其他交通方式的使用量预计将下降 4%–5%。